Mutations in numerous pathways need to accumulate for cancer to progress. Given the ability of transposons to cause mutation and the role of mutation in cancer, it seemed likely that transposons would play a role in cancer. A few years ago, Iskow and colleagues showed that transposons jump in some lung tumors, suggesting a link to cancer progression. They also showed that methylation levels are often lower in lung cancers. Methylation is important for transposons silencing, so they hypothesized that lowered methylation in cancer could lead to more transposon jumps. This would “destabilize” the genome, allowing more mutations to accumulate, and accelerating cancer progression.

However, very little evidence of this connection existed until recently. With the advent of high-throughput sequencing, it is becoming possible to examine changes in the genomes of cancer cells. Lee and colleagues report on one such study. They decided to look at the effect of retrotransposons by comparing the location of these jumping genes in normal and cancer cells. Retrotransposons copy their sequence from one location to another by going through an RNA intermediate that is read “backwards” from RNA to DNA.

In their study, they had to overcome a problem: because transposons are found throughout the genome and are mostly the same in different individuals, it is hard to figure out exactly where new transposons are located. To sort this out, they developed a bioinformatics tool that could align sequence to a reference genome and identify new transposon sequence associated with this sequence. They then used normal tissue and cancer tissue from the same individual to identify transposition events in cancer cells.

Interestingly, different cancer types had different numbers of transposon jumps. Brain and blood cancers did not have many transposon-induced mutations, while epithelial cancers had frequent insertions. These jumping-gene insertions are probably important for cancer, as many of the insertions occur within genes known to affect cancer biology.

If these jumping genes cause mutations and promote cancer, why are they there? It’s still an area of contention, but all that jumping around helps provide diversity in our genomes. Sometimes that will prove to be bad, but it also allows natural selection to act on the diversity, allowing new, helpful innovations in our DNA power evolution.

Iskow el al, 2010. Natural mutagenesis of human genomes by endogenous retrotransposons. Cell. 141(7):1253-61.

Lee et. al, 2012. Landscape of Somatic Retrotransposition in Human Cancers. Science. 337(6097): 967-971.

Tumor cells are a mosaic of different cell types

For the last decade or so, progressive cancer treatments involved taking samples of tumors, testing the cells to determine the genetic makeup, and then prescribing medicines targeted to specific mutations. There are many benefits to this approach, but it doesn’t always work.

It turns out that tumors aren’t uniform; they are mosaics of cells that can be genetically very different. A recent paper in the New England Journal of Medicine showed that a cell in one area may not be the same as a call in another area (a phenomenon called “intratumor heterogeneity”). So a treatment based on a sample from one area may not work for the whole tumor. Some tumor cells may be resistant to the drug so the cancer persists, or even grows.

In this British pilot study, cells from 9 different locations within a primary kidney tumor, and several metastatic tumors, were analyzed using next generation DNA sequencing. Only 34% of the 118 mutations identified were present in all the samples, and several of the major cancer genes were mutated in different ways in different locations. This turned traditional ideas about cancer cells being “clones” of a single, mutated cell on its head.

Previously, it was thought that a tumor develops when a single cell accumulates sufficient mutations over time that eventually lead to it dividing uncontrollably. Therefore if you could find the original mutation, and target treatment to that, then every cell would react to the treatment. But if the tumor is made up of a mosaic of cells, then they could all react differently to the drug. The researchers then created a phylogenetic “tree,” identifying which cells were more persistent, being in the trunk of the tree. They proposed that if those cells were receptive to a targeted medicine, the treatment might be more effective; if not, less so.

Although this study only involved four patients, the results provide a new way of thinking for researchers and clinicians. If we remove the presumption that all tumor cells are identical, we open the way for more creative thinking about how to tackle the problem.